Findings

Demographic Overview
Characteristics of Illegal Households. Table 1 reports demographic information for households headed by illegal immigrants and all other households. Not surprisingly, it shows that on average households headed by illegal aliens have much lower average incomes and are somewhat larger in size than the average household in America. The lower income reflects not simply their legal status, but more importantly the fact that such a large share of illegals have little formal education. An estimated two-thirds of illegals who are household heads lack a high school diploma. This is important because it means that even if the illegal aliens were legal residents, their income still would be dramatically lower than the rest of the population. There is no single better predictor of income in the modern American economy than one's education level. As a result, a large share of illegals are likely to remain poor even if given legal status. Table 1 also shows the share of households in which at least one person works. A much larger share of illegal households had at least one person working in 2002 than non-illegal households. Thus, any costs associated with illegal aliens do not reflect low rates of employment.

Illegal Household Use of Services. The lower portion of Table 1 shows the percentage of illegal households receiving Social Security and means-tested programs. There are very large differences in program usage between illegals and the rest of the population. Just as important, illegals' use of different types of programs vary enormously. Only a tiny fraction of illegal households use Social Security or cash welfare programs compared to other households. One source of public dissatisfaction with illegal immigration is that some Americans believe that many illegal aliens are getting cash welfare payments. Table 1 shows that this is not the case. However, the share using Medicaid and food assistance welfare programs is quite high and substantially more than the share of non-illegal households.

It must be remembered that, for the most part, illegal households using programs like free school lunch or Medicaid are receiving these benefits on behalf of U.S.-born children, who under current law are awarded citizenship at birth. Of course, the costs of providing services to these children are very real for taxpayers and result from illegals having been allowed to enter and stay in the country. And having the federal government feed or provide medical care to their children is an enormous benefit to illegal aliens. Thus, in considering the consequences for public coffers, counting the costs of these programs is necessary, otherwise one would gain a very false sense of illegal immigration's present costs. Nonetheless, the fact that it is the U.S.-born children receiving the benefits is still important, because it means that barring illegals from using programs would not significantly reduce costs. Their citizen children would continue to receive them. On the other hand, if the illegal families were made to return home, the costs would be eliminated.

Estimated Tax Payments
Payroll Taxes. Table 2 shows a breakdown of the estimated tax payments and services used by illegal alien-headed households. (More details about illegals' payment of specific taxes and use of specific programs use can be found in the Appendix on p. 39.) In terms of tax payments, the table shows very large differences between illegal households and other residents. The largest difference is in federal income taxes. Illegal households pay only about one-fifth as much as other households. This is not surprising given their much lower incomes and larger family size. By design, households with modest incomes and large size are supposed to pay relatively little in taxes. This, coupled with the fact that only a little over half of illegals make payroll contributions, is the reason their payments are very low relative to other taxpayers. For taxes other than income tax, the difference between illegal households and all others is not quite as large. Because Medicare and unemployment are more regressive in nature than federal income tax, the contribution of illegals for these two taxes is about 37 percent that of other households' contributions. And for Social Security, which is even more regressive, illegals pay 40 percent of the average household's contribution.

It must be remembered that tax payments in the table are based on the assumption that only 55 percent of illegals pay payroll taxes, comprised of income tax, Social Security, Medicare, and unemployment insurance. If all of their income were subjected to taxation, illegals' tax payments would rise significantly. Of course, if they paid all of their payroll tax liability, this would imply that they have legal status, which would also dramatically increase use of public services. This issue will be discussed later in this report.

Excise and Other Taxes. For excise and estate taxes, which are the most regressive, illegals pay 55 percent as much as other tax payers. In Table 2, estate and excise taxes are grouped together. Because illegals are very young on average, they pay very little in estate taxes. If only excise taxes are considered, the average tax payment of illegal households is 69 percent that of all households. Thus, there are significant differences in the relative size of payments illegals make to the various programs. The more regressive the tax, the closer the relative payments of illegal households are to the rest of the population. While the tax payments made by illegal aliens are much smaller on average than those of other households, illegals still do pay billions of dollars in taxes to Washington. In 2002, illegal households paid a total of nearly $16 billion to the federal government. The far right column in Table 2 shows that illegal alien tax payments constitute about 0.9 percent of all taxes collected. Because persons in illegal households constitute 3.6 percent of the nation's total population, their tax payments are clearly less than their representation in the population as a whole. This fact by itself does not mean that they create a fiscal deficit, because the net effect of illegal households on public coffers also depends on their use of public services, which is discussed below.

Costs by Household
Social Security and Medicare. The lower half of Table 2 reports the estimated costs illegals impose on public coffers. The table shows that, in general, illegal households use much less in almost every type of service. In the case of Social Security and Medicare, illegal households use about one-twentieth as much as other households. And they account for less than two-tenths of 1 percent of the total cost of these very large programs. Moreover, it is also clear that illegals pay substantially more in Social Security and Medicare than they use, creating a net benefit for these two programs of over $1,800 a year per illegal alien household. This calculation actually understates the benefit illegals create for the trust funds of these two programs because Table 2 includes costs for Medicare part B, which is paid for by general funds. If only Medicare part A (the part of Medicare covered by the trust fund) and Social Security are considered, illegal households create a net benefit well in excess of $7 billion dollars a year for the trust funds of these two programs.

Welfare Programs. Table 2 shows that illegal households receive much less in cash assistance welfare programs. As already discussed, persons in illegal households comprise 3.6 percent of the total population, but their use of cash welfare accounts for 0.3 percent of costs for these programs. For food assistance programs, however, illegal households actually receive more of this type of program than non-illegal households, accounting for 5.6 percent of federal costs for these programs. This is mainly due to heavy use of the WIC and school lunch programs. For Medicaid, illegals receive less than other households, but their use of this very expensive program is still significant. It's worth noting that although Table 1 showed a larger share of illegal households using Medicaid, figures in Table 2 show that on average they receive a lower payment. This mainly reflects the fact that it is typically only the U.S.-born children in the illegal households who are on Medicaid, while in other households with low incomes both parents and children can qualify for the program. Table 2 also shows that illegal use is much lower for all other welfare programs. But the table still shows that illegal households do use these programs to some extent.

Other Transfers to Households. For other transfers to households, the $442 illegal households are estimated to receive is about half of what non-illegal households get and certainly much less than their share of the total population. The programs included under this category are listed at the bottom of the table. Eligibility and use vary a great deal. For example, illegal aliens who work in the United States illegally are explicitly allowed to receive the Additional Child Tax Credit, which pays out a total of $5 billion a year to low-income workers with children. On the other hand, programs such as student loans can only be used by the tiny number of citizen children in illegal households who are enrolled in college. As is the case with most means-tested programs, illegals use considerably less than other households, but their use is not zero. Overall, illegal households account for less than 2 percent of the costs of these programs.

Prisons, Schools, the Uninsured, and Immigration. There are four areas where the estimated costs illegal households impose are much larger than for other households--treatment for the uninsured, federal aid to schools, federal prisons/courts, and the immigration system. Figures for the uninsured simply reflect the fact that such a large share of illegal aliens and their children lack health insurance. With more than half of persons in illegal households lacking health coverage, illegal households account for a very large share of the costs of treating the uninsured. As for schools, although they are primarily paid for by state and local governments, the federal government now provides more than $28 billion for primary and secondary education. Moreover, Washington gives schools some assistance in paying for children with limited English and for the children of agricultural workers. Not surprisingly, illegal households account for a disproportionate share of these programs. Illegal households impose very significant costs on the federal education budget, however, mainly because illegal households have more school-age children on average. The costs for the federal prison and court system are also significant because, although persons in illegal households account for about 3.6 percent of the nation's total population, illegals now account for almost one-fifth of those in federal prison and others processed by the federal courts. Thus, they impose costs on that system that are disproportionally high relative to their share of the total population. This is also true for the immigration system. As indicated in the methodology section, the costs of the immigration system are assigned based on the distribution by household of post-1980 non-citizens. This probably understates the costs of illegals to the immigration system because enforcement alone, which is directed specifically at illegal aliens, accounts for two-thirds of the immigration budget. Nonetheless, it is certainly not surprising that illegal aliens account for a large share of the costs of the immigration system because so much of that system is focused specifically on them.

Balance of Tax and Cost
Illegals Create Large Net Costs. The bottom portion of Table 2 adds together the total tax payments and costs illegals impose on the federal budget. When defense spending is not considered, illegal households are estimated to impose costs on the federal treasury of $6,949 a year or 58 percent of what other households received. When defense spending is included, their costs are only 46 percent those of other households. However, they pay only 28 percent as much in taxes as non-illegal households. As a result, the estimated net cost per illegal household was $2,736. Whether one sees this fiscal deficit as resulting from low tax payments or heavy use of services is a matter of perspective. As already discussed, illegal households comprise 3.6 percent of the total population, but as Table 2 shows they account for an estimated 0.9 percent of taxes paid and 1.4 percent of costs. Thus, both their payments and costs are significantly less than their share of the total population. Since they use so much less in federal services than other households, it probably makes the most sense to see the fiscal deficit as resulting from low tax payments rather than heavy use of public services.

Total Deficit Created by Illegals. If the estimated net fiscal drain of $2,736 a year that each illegal household imposes on the federal treasury is multiplied by the nearly three million illegal households, the total cost comes to $10.4 billion a year. Whether one considers this to be a large sum or not is, of course, a matter of perspective. But, this figure is unambiguously negative and certainly not trivial. It is also worth remembering that these figures are only for the federal government and do not include any costs at the state or local level, where the impact is likely to be significant.

The Fiscal Implications of Amnesty
So far we have only considered the current fiscal impact of illegal alien households. In the following section we run two different simulations to estimate what would happen if illegal aliens in the United States were legalized. We assume that any amnesty that passes Congress will have Legal Permanent Residence (LPR, colloquilly known as a "green card") as a component. It is true that President Bush's amnesty proposal in January 2004 envisioned temporary worker status for illegal aliens. At present, however, every major legalization bill in Congress provides illegals with LPR status at some point in the process. Moreover, because Republicans are divided between those favoring enforcement of immigration laws and those who want an amnesty, any legalization must have significant Democratic support to pass. But Democrats have made clear than they can only support an amnesty that gives permanent residence. While a two-step legalization -- one that grants temporary status before permanent residence -- is certainly possible, such a system would still result in permanent residence. Politically, it seems almost certain that any amnesty that actually passes Congress will award LPR status to illegals. But even if one makes the very unlikely assumption that an amnesty will be a pure guestworker program, the net fiscal deficit imposed by illegals indicates that unskilled workers who are not permanent residents still create large fiscal costs. As we have seen, this is partly because of their U.S.-born children, partly because like all people they necessarily place some demands on government, and partly because their low-income results in very low tax payments. All these things would still be true of unskilled guest workers.

General Impact of Amnesty. It is important to consider the likely outcomes of any amnesty: First, there should be a significant increase in tax compliance. (In the simulations below we assume that compliance rises from 55 percent to 100 percent.) Second, the average income of illegals should rise, as they would be freer to make decisions about employment and less likely to be exploited by employers. Third, use of public services will increase as the now-legal immigrants are eligible for many services from which they were barred as illegals. The actual size of these changes and their impact on the fiscal bottom line are explored below. It is also very important to realize that, if legalized, illegal aliens would not simply become just like legal immigrants in general because illegals are much less educated on average than legal immigrants. To run our legalization simulations we use the characteristics of illegals who are household heads and then assume that if they were legalized, they would pay taxes and use services like legal immigrant households headed by persons with the same characteristics.

Simulation One. We first report the education levels and country of birth of illegal households based on the education and country of the household head. Figure 1 reports the Mexican and non-Mexican share by education level of the illegal alien population. We divide illegals by their education level because, as already discussed, the single most important determinant of one's income, and thus service use and tax payments, is education. This was one of the most important conclusions of the aforementioned NRC study. We further divide them by whether they are Mexican because all research, including this report, indicates that about 60 percent of illegals are from that country. Moreover, the NRC study found significant differences between households headed by Latin American immigrants and those from the rest of the world. Figure 1 reports the distribution of illegal alien household heads by their educational attainment and if they are Mexican. We then assume that if illegals were legalized they would use services and pay taxes like all legal immigrants with the same characteristics.³²

Simulation Two. In the second simulation we again divide the illegal population by education and whether they are Mexican, but this time we assume that they would pay taxes and use services like legal immigrants who arrived in 1986 or after. That is, we again use the results from Figure 1 and assume that if legalized, they would use services and pay taxes like legal immigrants who have the same characteristics, but have arrived since 1986. The reason we assume that they would be like post-1986 legal immigrants is that illegals themselves are almost all post-1986 arrivals, the year the last amnesty for illegals was passed. Illegals are much more like recently arrived legal immigrants then they are like legal immigrants in general, who are older and have higher income but also use programs like Social Security. In addition, we exclude persons from the main refugee-sending countries because it is well established that refugees have the highest rates of public benefit receipt of any group of immigrants. These countries include: Poland, the former Yugoslavia, the Former Soviet Union, Afghanistan, Cambodia, Iraq, Laos, Vietnam, Nicaragua, Cuba, and Ethiopia. This second simulation is probably the most plausible.

Fiscal Impact of Legal Immigrants. Table 3 shows estimated federal taxes paid and services used in 2002 for legal immigrant households by education level and whether a person is from Mexico. The left side of the table is the basis for Simulation 1 and shows all legal immigrants regardless of when they arrived; the right side of the table shows the same figures for post-1986 non-refugees. One interesting finding of the table is that estimated tax payments are higher for all categories of legal immigrant households than the roughly $4,200 paid by illegal households in Table 2. The only exception is for post-1986 Mexicans without a high school education. Thus, those who contend that legalization would increase tax revenues are probably correct. Almost every category of legal immigrant pays more in taxes than do illegal households. Unfortunately, the table also shows that, in every case, total federal costs are also higher than the roughly $6,900 a year reported for illegal households in Table 2. Thus, those who are concerned that legalization would increase costs are also almost certainly correct.

(Click on Table for a larger version)

As expected, Table 3 shows that there is very wide variation in public service use and tax payments between groups. Overall service use and tax payments by household closely correlates with education levels of household heads. In both simulations, those with more than a high school degree are a large net fiscal benefit to the federal government, while those with only a high school education or less are a net fiscal drain. This is true whether the legal immigrant is from Mexico or not. It is also true whether one considers all legal immigrants or only post-1986 non-refugees. The difference between immigrants by education is truly enormous. For example, looking at households headed by post-1986 non-refugees, a legal immigrant without a high school degree creates a fiscal deficit on the federal government of more than $15,000 a year if he is Mexican and almost $11,000 if he is from the rest of the world. Conversely, for those with more than a high school degree, the benefit is over $5,000 a year if they are Mexican and more than $12,000 for non-Mexicans. The same pattern holds when all immigrant households are considered, regardless of year of arrival. Without question, the education level of legal immigrants is a key determinant of their fiscal impact. However, being from Mexico seems to matter as well.

Legalization Would Dramatically Increase Costs. Using the education levels and the share that is Mexican found in Figure 1 and combining them with the results from Table 3, we can then estimate the likely impact of legalization. Table 4 reports the estimated federal taxes paid and services used by legalized illegal alien households assuming they would pay taxes and use services like households headed by legal immigrants with the same characteristics. The first simulation assumes that illegals would pay federal taxes and use services like all legal immigrants with the same education level, regardless of when they arrived, while the second simulation assumes that they would have the impact of post-1986 legal non-refugee immigrants. Both simulations show that legalization would increase the net fiscal costs dramatically. Simulation 1 shows that the net fiscal costs to the federal government would increase from $2,736 to $6,022 per household. Simulation 2 shows the net fiscal cost would be even larger, increasing to $7,668 per household. Although net costs rise, estimated tax payments increase dramatically with legalization. They more than double in Simulation 1 and increase by 77 percent in Simulation 2. However, costs rise significantly as well. In both cases costs more than double, creating an average net fiscal deficit per household that is significantly more than that estimated for illegal alien households.

Interestingly, total costs per household are roughly the same in both simulations. In Simulation 1, legal immigrants are making significantly more use of Social Security and Medicare, while in Simulation 2 use of welfare, education, and other transfers to households are much higher. This reflects the different age structure of post-1986 immigrants compared to all immigrants. In effect, one can see Simulation 2 as the more immediate impact of a legalization, while Simulation 1 reflects the likely fiscal impact of legalized aliens in the long term. Either way, the results indicate that legalization would cause the net fiscal burden on taxpayers to increase substantially. It should be noted that, although legalization would substantially increase costs, this does not mean that legal immigrants overall are net drain on the federal government. It also worth noting that if only the relatively small share of illegal aliens with more than a high school degree were legalized, then there would be no fiscal deficit. But since so many illegals have only a high school degree or less, legalizing all illegals would create a large fiscal deficit.

Would Welfare Reform Hold Down Amnesty Costs? In 1996, Congress passed the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA), which barred some legal immigrants who arrived after 1996 from using certain welfare programs. Since many legal immigrants used in Simulation 2 to estimate the costs of amnesty arrived before 1996, perhaps our estimates overstate the costs, at least with regard to welfare programs. Table 5 provides a detailed breakdown by program for all of the costs estimated in this study. The table shows that legalization under Simulation 2 would increase costs by a total of $8,173 per household. (The same total found in Table 4) Of that increase, $5,588, or about 68 percent, is the result of increased costs due to cash payments from TANF, SSI, and in-kind benefits received from food stamps and Medicaid, the four main programs covered by PRWORA. While certainly a large share of the costs, it still leaves an increase of $2,585 per household in non-PRWORA covered programs.

More importantly, analysis of the CPS shows that households headed by recently arrived legal immigrants actually do make significant use of these welfare programs. Table 6 reports estimated use of these four programs for households headed by legal immigrants from Mexico and the rest of the world, excluding refugee countries. The table shows that, in 2002, households headed by a legal immigrant from Mexico who arrived in 1996 or later and who lacks a high degree received $7,900 on average from these four programs and those with only a high school degree received $3,817.³³ For post-1996 households headed by legal non-refugees from countries other than Mexico without a high school degree, the figure is $5,464 per household, and for those with only a high school degree the figure is $2,462. By contrast, the figure for illegal aliens is only $860 per household.

Table 6 also shows that if we assume that illegal households, once legalized, would use welfare like non-refugee legal immigrants who arrived in 1996 or later with the same education, then welfare costs would be $5,430 per household, very similar to the $5,588 in Simulation 2. It must be remembered that, as originally shown in Table 3, legal immigrants with little education make very extensive use of welfare programs because their incomes are very low and because they receive these programs on behalf of their U.S.-born children. This is especially true for Medicaid, by far the costliest program. Welfare reform has not changed these basic facts. Given this reality, it is very difficult to see how a dramatic rise in welfare costs could be avoided if there was a legalization.

Amnesty Would Increase Incomes. Although the net fiscal burden would increase dramatically as a result of a legalization, this does not mean that the income of illegals would remain unchanged. Table 4 shows a dramatic increase in payroll taxes and this partly reflects increases in household income that would occur if illegals were legalized. In fact, analysis shows that whether they are Mexican or not and no matter what their educational attainment, legalization would significantly increase the income of illegal aliens. Figure 2 reports the estimated income of legalized illegal households under Simulations 1 and 2. Like the estimates in Table 4, they are based on the assumption that, if legalized, illegal households would have average incomes like their legal counterparts controlling for whether the household head is Mexican and for education levels. Under Simulation 1, their income would be 29 percent higher, and under Simulation 2 incomes are 15 percent higher. Of course, the much higher incomes in Simulation 1 partly reflect the longer time that all immigrants have lived in the United States and not simply the fact that they have legal status. But Simulation 2 shows that even when analysis is confined to more recent arrivals, incomes still increase substantially. While our overall conclusion indicates that legalization would increase fiscal costs, perhaps the improvement in the income and economic well being of illegal households could be used to justify legalization. Certainly, Figure 2 indicates illegals would likely benefit significantly from obtaining legal status.

Comparisons to Other Studies
There is always some uncertainty associated with any fiscal estimate of immigrants, especially illegal aliens. However, the overall results of this study are consistent with other work that has looked at this question.

NRC Study. As mentioned at the beginning of this report, the 1997 study by the National Research Council (NRC) is probably the most sophisticated report ever to examine the impact of immigration on public coffers in the United States. While that report did not consider the legal status of immigrants, it did divide immigrants by education. The overall findings of that report showed that the education level of immigrants was critically important in determining their income, tax payments, service use, and resulting net fiscal impact. They found that during his lifetime an immigrant without a high school education imposed a net fiscal drain on taxpayers of $89,000; for those with only a high school degree it was $31,000. The NRC study also found that for those immigrants with more than a high school degree the fiscal benefit was $105,000. While we report estimates for only legal immigrants for a single year, our analysis by education found in Table 3 is strongly supported by the NRC's findings.

Urban Institute Study of New York State. One of the few studies to specifically estimate illegal immigrant tax payments at the federal level by household was done by the Urban Institute in 1998. That study used the CPS to estimate 1994 tax payments for illegal households in New York State, including some federal taxes. Although our study is designed to measure the fiscal impact of immigration at the national level, we ran separate estimates for federal tax payments for illegal-headed households in New York State so we could compare them to those of the Urban Institute. The Urban Institute only examined federal income tax, Social Security, and unemployment insurance, and we did the same. We estimated that the average illegal household in New York State paid $5,538 in these three taxes in 2002, compared to $4,568 estimated for tax year 1994 by the Urban Institute. Adjusted for inflation between 1994 and 2002, this is $5,463 or only 1.4 percent less than our estimate for 2002.³⁴ Of course, illegal immigration to New York State and the federal tax code has changed since 1994. Nonetheless, our average estimates are very similar to those developed by the Urban Institute when adjusted for inflation.

Treasury Department Report. The results of our study are also buttressed by an analysis of illegal alien tax returns done by the Inspector General's Office of the Department of Treasury in 2004.³⁵ That study found that 55 percent of illegal aliens who filed income tax returns using tax identification numbers had no federal income tax liability. While higher than the 45 percent estimated in this study, it must be remembered that this figure was only for those who filed tax returns using a tax identification number. These are all individuals who expected refunds, otherwise they would not have gone to the trouble of getting tax identification numbers and filing a return. But like our estimates, the Inspector General's report indicates that roughly half of illegals have no tax liability, reflecting their very low incomes and large number of dependents.

Return to Table of Contents

Continue to Conclusion

³² Although we included them in the costs for non-illegal households in Table 2, we exclude costs for the small number of federal programs that specifically aid refugees because if legalized, illegals should not access these programs.
 
 ³³ These recently arrived unskilled legal immigrants are relatively small in number and so the survey should be interpreted with some caution, but in the case of both Mexicans and non-Mexicans, use of welfare programs by recently arrived unskilled immigrants (excluding refugees) is quite high.
 
 ³⁴ Table 6 on page 103 of the Urban Institute study has the estimates by type of tax.
 
 ³⁵ The Inspector General's Office of the Department of the Treasury report # 2004-30-023 can be found at http://www.treas.gov/tigta/2004reports/200430023fr-redacted.html